Sorption of Pesticides and its Dependence on Soil Properties

Oct 28, 2014 - Sorption is one of the major processes that determine the fate, effects, efficacy ..... (42), in a most comprehensive analysis, assesse...
0 downloads 0 Views 1MB Size
Chapter 12

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Sorption of Pesticides and its Dependence on Soil Properties: Chemometrics Approach for Estimating Sorption Rai S. Kookana,*,1 Riaz Ahmad,2,3 and Annemieke Farenhorst4 CSIRO Land and Water, and University of Adelaide, PMB No. 2, Glen Osmond, SA 5064, Australia 2Department of Soil Science & Soil Water Conservation, Pir Mehr Ali Shah Arid Agriculture University, Rawalpindi, Punjab, Pakistan 3Punjab Agricultural Research Board, Lahore, Punjab, Pakistan 4Department of Soil Science, University of Manitoba, Winnipeg, MB R3T 2N2, Canada *E-mail: [email protected].

1

Sorption is one of the major processes that determine the fate, effects, efficacy and ecological risks of pesticides in terrestrial and aquatic environments. In this chapter we provide an overview of sorption and its dependence on soil properties. Soil solid phase consists of mineral and organic matter; both in strong association with each other make a contribution towards sorption of pesticides, depending on their relative abundance in soil/sediments, chemistry and the chemical nature of the pesticide molecule. We discuss the roles of organic matter and clay contents as well their chemistries in determining sorption of pesticides, and assess the partition theory in terms of its adequacy in describing the observed sorption behavior of pesticides in soil. The complex interactions and heterogeneities associated with the soil solid phase contribute to the large degree of variation in Koc, a parameter that is often used to extrapolate the pesticide sorption estimate, among soils. A more comprehensive approach incorporating soil organic matter as well as soil minerals (both contents as well as their chemistries) is therefore desirable to fully incorporate

© 2014 American Chemical Society In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

the role of soil surfaces in sorption of organic compounds. Therefore, we introduce some emerging approaches based on chemometrics and infrared spectroscopy that appear promising for comprehensive representations of the combined role of organic matter, mineralogy and other soil properties in estimating pesticide sorption in soil.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Introduction Sorption is one of the major processes that determine the fate, effects, and ecological risks of pesticides in terrestrial and aquatic environments. The sorption process commences as soon as a pesticide molecule comes in contact with soil/sediment and occurs mainly in the scale of hours. Other processes such as movement, persistence, efficacy, bioavailability, and ecotoxicity in the environment are often relatively slower and are affected by the sorption process. Therefore, the sorption parameter of a pesticide in soil is of fundamental importance in considering not only its effectiveness (and choosing appropriate application rate for a given soil type) but its potential risk to the environment. In this chapter, we provide an overview of sorption and its dependence on soil properties. We discuss the role of organic matter and clays, their contents and chemistries in determining sorption behavior of pesticides. We scrutinize the validity of the partition theory in describing sorption behavior of pesticides purely on the basis of soil organic carbon content (SOC) alone. We also introduce emerging approaches, especially those based on chemometrics and infrared spectroscopy, that appear promising for comprehensive representations of the combined role of soil organic matter (SOM) and mineralogical properties of soils.

Pesticide Sorption Processes in Soil and Sediments Uptake of organic molecules by a solid phase from a gaseous or liquid phase is generally referred to as adsorption. However, some organic compounds including non-ionic pesticides are also believed to exhibit a phase-partitioning process (absorption), akin to liquid-type interactions of two immiscible solvents. This is based on the assumption that the macromolecular SOM acts as an immiscible solvent leading to a partitioning process, similar to that between n-octanol and water. The octanol-water partitioning coefficient (Kow) is commonly used to represent the hydrophobicity of a compound. Since the two processes (adsorption/absorption) are difficult to distinguish experimentally in soils and sediments, the term ‘sorption’ is preferred, which makes no distinction between absorption, adsorption and precipitation of organic compounds (1). In addition to neutral compounds, many pesticides are ionic or inonizable in nature. Due to the presence of multiple functional groups, pesticide molecules can donate or accept protons depending on environmental conditions. The cationic, anionic or zwitterionic species are formed depending on the chemistry of the molecule and/or the pH of the surrounding environment. Sorption of pesticides 222 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

on a solid phase can involve a variety of mechanisms, ranging from London-van der Waals forces, hydrogen bonding, ligand exchange, ion exchange, cation-water bridging, covalent bonding and even physical trapping (1). Sorption from the perspective of sorbate as well as sorbent has been discussed by Pignatello (2), especially on organic carbon and black carbon in soils. The extent of sorption by a solid phase is determined by the nature and properties of both sorbate (e.g. a pesticide) and sorbent (soil or sediment) as well as ambient conditions during the equilibrium process (e.g. temperature, water content). Chemical characteristics of the aqueous phase which are important for sorption processes include pH, ionic strength, and redox status, all of which affect surface properties of sorbent, interactions between liquid and solid phases, and the ionization state and reactivity of the sorbate. Indeed, liquid-solid interface characteristics can have important impact on sorbate-sorbent interactions.

Important Soil Properties Affecting Sorption A number of soil properties including its chemistry, mineralogy, SOM, and pH, and environmental factors such as soil water content and temperature can influence pesticide sorption in soil. These soil properties are very important in determining the mechanisms of sorption and thereafter mobility of a pesticide in soil. Due to close association between the mineral and organic matter, it is often difficult to treat these as separate properties, in the context of sorption. Soils are inherently heterogeneous in nature and how variation of soil parameters influences the sorption interactions of pesticides is discussed below.

Organic Matter Soil organic matter has been recognized as one of the most relevant soil properties in regulating sorption of non-ionizable pesticides (3, 4). Strong correlation has been frequently observed between sorption of non-ionizable pesticides and the SOM content (5). However, ionizable pesticides such as weakly basic triazines can be sorbed to both SOM and clay minerals, and the sorption is pH dependent (4). For non-ionic pesticides, the humified materials in soil, due to their oxygen-containing functional groups such as –COOH, phenolic, aliphatic, enolic, –OH, and C=O, have often been found to be good sorbents (6). The content and the chemistry of SOM vary from soil to soil in terms of its polarity, elemental composition, aromaticity, condensation, and degree of diagenetic evolution from a loose polymer to condensed coal-like structures (7–10). Both the type and the age of SOM may affect sorption of pesticides. The type of vegetation from which the SOM originates affects its chemical composition and degree of decomposition (11). For example, humic materials of grassland soils are rich in humic acids, whereas fulvic acids dominate in forest soils (12). The O-alkyl C tends to decrease with decomposition, while the proportion of alkyl C tends to increase and aromatic C may increase or decrease, depending on the conditions (13–15). 223 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Soils around the world have been found to contain different amounts of black carbon such as charcoal (16), a highly aromatic condensed ring structure. In addition, biochar (charcoal produced specifically for agricultural use) is attracting increasing attention as a soil amendment to sequester carbon and enhance soil fertility (17). Black carbon or biochars have been noted to have very high affinity for organic compounds including pesticides (18–20). Yang and Sheng (18) reported that biochar produced from wheat and rice residues was 400-2500 times more effective than soil in sorbing diuron herbicide. The high specific surface area and aromatic nature of carbon in biochars may be responsible for their extraordinary sorption properties (2, 21). However, the capacity of biochars to sorb pesticides may depend on soil types. For example, Ahmad and colleagues (unpublished data), recently studied the sorption behavior of carbaryl in two contrasting soils from New Zealand previously amended with biochars produced from corn stover at 350 °C and 550 °C. They found that while the amendment of soils with biochar at an agronomic application rate of 7.2 t C ha-1 enhanced the sorption of carbaryl significantly; the magnitude of this effect varied considerably depending on soil type and the pyrolysis temperature used to produce the biochar.

Soil Mineralogy Clay minerals can sorb organic compounds that are cationic in nature (e.g. diquat and paraquat herbicides) or those containing polar functional groups (23–25). The clays in highly weathered soils are dominated by kaolinite (1:1 type of clay minerals) and have lower capacity for pesticide sorption than 2:1 type clay minerals such as montmorillonite and vermiculite (26). The high surface acidity of hydroxyinterlayerd clays and their complexes can increase the protonation of weakly-basic herbicides such as atrazine and, consequently their sorption by ionic bonds at low pH, and their sorption by physical forces (H-bonds) at near-neutral pH (22). Many soils, especially in the tropical regions, contain variable-charge minerals (e.g. iron oxides, gibbsite) with distinct physical and chemical properties (27). For example, Oxisols can carry net positive charge at ambient pH and can effectively sorb anionic pesticides (28).

Other Soil Properties Unlike hydrophobic non-ionic compounds, sorption of ionizable pesticides is highly sensitive to variations in the soil pH. This relationship mainly originates from different proportion of ionic and neutral form of the pesticide present at each pH level and also from the presence of surfaces with pH-dependent charges in soils. Soil pH regulates the electrostatic charge of soil colloids (organic matter and oxides) and protonation or chemical dissociation of pesticide molecules. Basic pesticide molecules become protonated at lower pH and therefore, more strongly sorbed to the soil colloids. On the contrary, the acidic pesticides ionize and become anions as pH increases (one or more pH units above the pKa) and exhibit lower sorption (28, 29). Soil pH varies markedly among soils of different regions. The soils in wet tropics are generally acidic, whereas those in arid regions 224 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

are highly alkaline leading to markedly different sorption behavior. This effect makes direct extrapolation of pesticide sorption data, particularly for ionizable compounds, between acidic and alkaline soils difficult. Soil temperature and water content are other important environmental parameters that may influence pesticide sorption. The effect of soil temperature on pesticide sorption has been noted to be highly variable. Valverde-Garcia et al. (30) and Khan et al. (31) noted that the elevation of temperature favoured sorption of pesticides in soils. They attributed this enhanced sorption to the increased number of active sites on humus. For 2,4-D sorption by 41 wetland sediments, Gaultier et al. (32) observed a significant increase in the 2,4-D sorption parameter when the experimental temperature was at 25°C rather than 5°C. By contrast, studies on sorption of cyanazine on different homoionic peats by Dios-Cancela et al. (33) showed decreasing sorption with increasing temperature, which was attributed to either a decrease in the attractive forces between the pesticide and the peat or a change in the solubility of the pesticide. In another study on sorption of atrazine on kaolinite and montmorillonite clays, Fruhstorfer et al. (34) postulated that since a rise in temperature causes an increase in the kinetic energy of the molecules with constant electrostatic attraction, this leads to a decrease in sorption of pesticides. Sorption may also be influenced by soil moisture status, but this effect is not well understood. Several studies by Koskinen and co-workers (e.g. (35, 36)) show that sorption of herbicides may increase with soil water content. Roy et al. (37) reported significant effects of water content on the sorption of five fungicides and these effects were found to be dependent on the properties of the chemicals and the sorption mechanism. Berglöf et al. (35) proposed that at higher water contents, more pesticide solution is in contact with a large surface area of soil particles, which facilitates the accessibility of the pesticide to sorption sites. Methods, such as the one proposed by Ahmad et al. (38), may help better understand the role of soil water content in sorption of pesticides.

Conventional Approaches of Estimating Sorption of Pesticides in Soils Pedotransfer Functions Pedotransfer functions based on soil properties have been developed to describe the sorption of pesticides in soils. For example, Weber et al. (39) correlated the literature reported pesticide Kd values for 57 pesticides from different chemical classes with the key soil properties (such as SOM, clay content and /or soil pH) and developed equations for estimation of Kd values for improved soil mobility predictions. Ahmad and Rahman (40) carried out a study examining the effect of soil properties mentioned above on sorption of two commonly used herbicides, atrazine (weak base) and imazethapyr (weak acid), in 101 allophonic and non-allophanic soils of New Zealand. A wide variation in the sorption affinities of the soils was found, as the Kd values of atrazine and imazethapyr ranged from 0.7 to 52.1 L kg-1 and 0.1 to 11.3 L kg-1, respectively. For atrazine, the sorption affinities for the allophanic set of soils (mean Kd, 8.5 L kg-1) were 225 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

greater than for the non-allophanic set of soils (mean Kd, 7.5 L kg-1). However, no effect of allophanic status was found on imazethapyr sorption. None of the measured soil properties could alone explain adequately the sorption behavior of the herbicides. Multiple regression analyses revealed that SOC content, pH and, to a lesser extent sand were the primary soil properties. The authors compared their calculations with Weber et al. (39) models and found limited congruence between the two (as shown in Figure 1). The above examples show that while pesticide sorption in soils can be correlated well to soil properties, it may not be easy to transfer the functions based on soils from one region to another.

Figure 1. Relationships between measured Kd values of atrazine and imazethapyr and corresponding Kd values calculated from Pesticides Properties Database (42) and Weber’s models (39). Reprinted with permission from Ahmad and Rahman, (40). Copyright (2009) American Chemical Society.

Quantitative Structure-Activity Relationship (QSAR) There are several indirect methods of estimating sorption based on chemical characteristics of compounds, ranging from solubility, Kow, retention factor in reverse-phase liquid chromatography and molecular structure. So called Quantitative Structure-Activity Relationship (QSAR) based approaches rely on molecular properties, such as molecular size descriptors, solvation energy relationships, molecular connectivity indices and other properties. These methods 226 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

have been discussed by Wauchope et al. (42) and by others cited in their review. Another chapter in this book (43) describes the QSAR approach in detail and applies QSAR for pesticide sorption. The appropriateness of the choice of a particular QSAR approach depends on the chemical structure and properties of the molecule (44) and their utility on the objective of the exercise (42).

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Partitioning Theory (Koc Model) Partitioning processes on SOM from aqueous systems have been widely accepted as a major contributing factor, especially for non-ionic pesticides (45). Consequently, organic carbon content of soil is often used to normalise the sorption coefficient (Kd) of pesticides to obtain Koc, such that Koc = Kd /fOC. where fOC is the fraction of organic carbon in soil. Assuming that Koc is a constant (i.e. sorption per unit carbon in various soils is the same), one can extrapolate Koc value from one soil to other. This approach has been widely accepted and used in various pesticide fate models. However, partitioning is a historic paradigm for sorption on SOM for non-polar solutes and, in most cases, it is a simplification for the purposes of modelling rather than driven by the evidence from sorption studies (2). A large body of evidence in literature shows that Koc, despite the users’ tendency to assume this as a constant, is a highly variable parameter. As early as in 1972, Hamakar and Thompson (46) and subsequently, Minglegrin and Gerstl (47) and Gerstl (48) highlighted the variability associated with Koc (KOM based on bulk SOM was used in some of these studies). More recently, Ahmad et al. (49) measured sorption of two pesticides (carbaryl and phosalone) on 48 soils collected from tropical and temperate regions involving three countries (Australia, Pakistan and the U.K.). They found only about half the variance (46-53 %) in Kd was explainable from the foc alone. Similarly, using 101 soils from southern Australia, Forouzangohar et al. (50) noted that while sorption of diuron showed dependence on SOM, only 42% of the variation in Kd was explained by foc. The sorption data reported by Ahmad et al. (49) for carbaryl measured in soils from various countries have been presented in Figure 2. The figure demonstrates that contrary to assumption of Koc being constant, the parameter shows strong dependency on the chemistry of SOC rather than its concentration in soil. Wauchope et al. (42), in a most comprehensive analysis, assessed thousands of Koc values of a large number of pesticides from a database (41) and identified the extent of variations in the parameter and factors contributing to this. They noted that typical coefficients of variation (CV) among reported values of Koc were in the range of 40-60%. The ratios between maximum and minimum values reported for the same compound from a single database were found to vary from 3 to 10 (Figure 3). The variation was found to be independent of the magnitude of Koc (i.e. low or high Koc reflecting hydrophilic or hydrophobic compounds). The authors ascribed about half of the observed variability in Koc to experimental errors plus the variability in measuring SOM and the remainder to the variability in organic matter itself, such as its chemistry. However, we believe that soil minerals (organomineral interactions) may also have contributed to the observed variability in Koc and should be taken into considerations. 227 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Figure 2. Dependence of sorption (Koc) of carbaryl insecticide in soils on the chemistry of soil organic carbon. Reprinted with permission from Ahmad et al. (49). Copyright (2001) American Chemical Society.

Disproportionate Reliance on Soil Organic Matter Content in Prediction of Sorption As discussed above, it is generally assumed that the sorption behavior of organic compounds is largely moderated by the content of organic matter in soils/ sediments, and the role of clay minerals is generally ignored, especially during risk assessment (51). This is despite the fact that often soils are much richer in the mineral matter, i.e. its content in soils is often an order of magnitude higher than that of organic matter. Besides, numerous studies over the years have demonstrated that clay minerals, especially expandable smectites, do have a strong affinity for a variety of organic compounds such as pesticides and nitroaromatic compounds. For example, Laird et al. (52) studied sorption of atrazine on soil smectite clays and showed that relatively pure smectites were quite effective sorbents of the herbicide, depending on their compositions. Similarly, clay minerals were found to be good sorbents for nitroaromatic compounds (e.g. (53)). Despite the large body of literature, the role of clays in determining the sorption behavior of organic compounds in soils/sediments, and consequently 228 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

their ecological risks, is still not fully appreciated and incorporated in the risk assessment models. Soil organic matter and clay minerals are often closely associated and the type of precursors of humic substances and chelating organic acids associated with hydroxyinterlayered clays also have a pronounced impact on the sorption of pesticides by clays (22). Hence, the impact of clay minerals on pesticide sorption needs to be studied by considering this association.

Figure 3. The observed values (A) and the variations in Koc values (B) as indicated by the ratio of maximum to minimum values for a range of pesticide active ingredients in soils. Reproduced with permission from Wauchope et al. (42). Copyright (2002) Society of Chemical Industry). 229 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

One of the major developments that led to a disproportionate dominance of the role of organic matter in sorption of organic compounds in soils is the introduction of the “partition theory” (45), mentioned earlier in this paper. This theory allowed an easy extrapolation of sorption measured on one soil to another based on its carbon content alone, using the Koc parameter. The simplicity of the concept led to models which could account for sorption of organic compounds in soil/sediments by a universal parameter Koc, applicable to all soils/sediments. However, as discussed earlier, this concept is a simplification based on the sorption behavior of non-ionic organic compounds, while a substantial proportion of pesticides are ionic or ionizable (weakly basic, acidic) in nature. For example, the most used herbicide glyphosate is a zwitterion, the bypridilium herbicides (paraquat and diquat) are cationic in nature and the triazines (e.g. atrazine) are weakly basic compounds. Indeed the low ecological risk profiles of most commonly used herbicides glyphosate, paraquat and diquat are due to their strong adsorption on clay minerals. But for the ability of soil clays to “detoxify” them rapidly, the environmental safety of these herbicides may not have been the same. Furthermore, the measured sorption isotherms of organic compounds in soils/ sediments are more commonly non-linear in nature, contrary to the assumption of the partition theory (2). The non-linearity of sorption isotherms is indicative of complex organo-mineral interactions reflecting the role of chemistry of organic matter and/or clays in determining the sorption behavior of organic compounds in soils. As discussed above, our studies (49, 54, 55) and those by many others (42, 56–58) observed that the organic matter content alone fails to adequately describe the sorption behavior of a wide range of pesticides in soils. Due to strong association between the two, the minerals and organic matter in soils are essentially inseparable (59) and organo-mineral interactions in soil may play a major role in sorption of organic compounds including pesticides (57). A more comprehensive approach incorporating soil organic as well as mineral matter (both contents as well as their chemistries) is therefore desirable to fully incorporate the role of soil surfaces in sorption of organic compounds.

Recent Advances in Estimating Pesticide Sorption Based on Integrated Soil Properties The recent developments in application of infrared spectroscopy (IR) in characterization of soils and sediments open a window of opportunity to estimate the sorption of organic compounds based on the overall contributions of mineral as well as organic matter in soils. Chemometric approaches, especially partial least squares (PLS) regression, with infrared spectroscopy have been found to be particularly effective in characterizing soils in terms of a range of soil properties, such as organic carbon content, clay content, cation exchange capacity etc. (60). Application of a couple of these approaches in estimating the sorption of pesticides has been discussed below. 230 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Prediction of Soil Properties by Infrared Spectroscopy

Several statistical measures can be used to evaluate how well IR predicts soil properties. The coefficient of determination (R2) and the root mean square error of prediction (RMSEP) or the standard error of prediction (SEP) are the most consistent parameters reported in studies. Using the SEP and bias separately is preferred to using RMSEP which integrates the bias in its value (61). The ratio of prediction to deviation (RPD) and the ratio of the SEP to the range (RER) are also commonly used to evaluate the accuracy of IR measurements (62). In general, it is understood that a larger R2, smaller SEP or RMSEP, and larger RPD indicate a more successful application of IR when measuring soil properties (63). While Dunn et al. (64) suggested 3 categories for RPD as poor, 2.0 for soil analysis. Malley et al. (2004) proposed that values of RPD >4 be judged as indicating excellent calibrations, >3 as successful, and between 2.25 to 3.0 as moderately successful. Near-infrared spectroscopy (NIR) is a technique that has long been recognized as having practical applications in agricultural and food industries. Its potential in soil research has also been demonstrated. For example, it has been used to successfully predict total carbon, or fractions of carbon in soils obtained from regions in Canada (62, 65, 67), United States (63, 67, 68), Australia (64), Uruguay (69), and Germany, Norway, and Denmark (70). NIR has also been used to predict soil textural characteristics (sand, silt or clay), pH, cation exchange capacity (CEC), CaCO3 content, and the total or exchangeable macronutrients (i.e., N, K, Ca, Mg, and P), micronutrients (Fe, Mn, Zn, Cu, Mo, and Ni), and metals (Pd, Co, Cd, Ag, V, Hg, Cr) (65, 66, 69, 71). The accuracy of these predictions ranged from less than successful to excellent calibrations. For the same parameter, one data set of soil samples may show successful calibrations (e.g. CEC shows R2 = 0.89 and RPD = 3.0 for Duck Mountain Provincial Park samples [n=108] in Malley et al. (71), while another data set of soil samples demonstrates poor calibrations (e.g. CEC shows R2 = 0.51 and RPD = 1.4 for agro-Manitoba samples [n=1,000] in Malley et al. (71). When small (e.g. n=100) or large (e.g. n=1,000) data sets are used, moderately to excellent NIR calibrations are common when measuring SOC and clay (63, 71, 72). Similarly, mid-infrared spectroscopy (MIR) combined with PLS regression approach has also been successfully used to estimate the content of soil major oxides, pH, sum of cations, Ca, Mg, K, Na, total N, clay content, and organic carbon content (60, 73). The technique has been particularly successful in predicting the SOC. Furthermore, both NIR and MIR techniques have also been applied to predict the chemistry of organic carbon. For example, Cozzolino and Moron (69) and Zimmermann et al. (74) showed that both MIR and NIR not only can estimate SOC in bulk soils but also in different particle-size fractions. Janik et al. (75) and Bornemann et al. (76) demonstrated that MIR could also be used for effective estimations of black carbon in soils. They employed MIR-PLS approach in analyzing the foc and black carbon contents in 309 soils collected from different depths in soils from America, Europe and Asia. In this study benzene polycarboxylic acid (BPCAs) was used as a specific marker of black carbon. The 231 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

prediction of BPCA carbon in the full set of soils (n=309) resulted in a R2 of 0.84 and an RPD close to 2.5, both indicating a secure predictive capability of the approach.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Prediction of Pesticide Sorption in Soils by Near- and Mid-Infrared Spectroscopy and Chemometrics Bengtsson et al. (77) were the first to demonstrate that NIR could be used to determine sorption coefficients (Kd) in soils. Their reference data consisted of Kd values of lindane (an organochlorine insecticide) and linuron (a substituted urea herbicide) determined for 27 soil samples by batch equilibrium experiments. Spectral data was collected between 1100 to 2500 nm at 2 nm intervals using a NIR 4600 scanning monochromator. The reference and NIR data demonstrated good agreement with R2 being 0.85 for lindane and 0.84 for linuron. Both pesticides showed good correlations that were even greater than the reported linear relation between SOC and lindane (R2=0.82) or linuron (R2=0.74) Kd values. Both lindane and linuron are neutral pesticides whose sorption in soil linearly increases with increasing SOC (77). Farenhorst and collaborators further investigated the feasibility of NIR as a tool for estimating Kd values for pesticides and related compounds (Singh et al. (66, 78)). Their study included the herbicides 2,4-D (a weakly-acidic), atrazine (a weakly-basic), and glyphosate (a zwitterion), as well as a natural steroid hormone, 17 β-estradiol. They determined the Kd values of these compounds in 609 horizons collected from 70 soil profiles in two irregular undulating to hummocky agricultural landscapes in the Canadian Prairies. Spectral data were obtained by scanning the 609 soil samples in 5 cm diameter glass Petri dishes to a 45VISNIR Zeiss Corona spectrometer using a wavelength range 380−1690 at 2 nm intervals, and to a Foss NIR Systems 6500 spectrometer with a wavelength range 1100−2500 at 2 nm intervals. The noisy region (380−700 nm) of the Corona spectra was removed so that, for the Corona, there was a wavelength range of 700−1690 nm. For both the Corona and Foss spectrometers, excellent to moderately successful calibrations were found for SOC and Kd values of 2,4-D, atrazine and 17 β-estradiol, with R2 ranging from 0.80 to 0.96, and the RPD ranging from 2.22 to 5.34 (Table 1). In contrast, the calibrations for glyphosate, pH and soil textural characteristics were largely unsuccessful in both landscapes and with both spectrometers (Table 1). These and the results of Bengtsson et al. (77) suggest that for herbicides such as atrazine, 2,4-D and linuron, whose soil retention is largely controlled by SOC, NIR is a good tool for quantifying sorption coefficients in soil, but it has limited value for measuring sorption coefficients for glyphosate, which is not significantly correlated with SOC (79) or may even decrease with increasing SOC (80).

232 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

233

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Table 1. Calibration Results of 2,4-D, Atrazine, Glyphosate and 17β-Estradiol Kd Values and Selected Soil Properties in Soil Samples Presented in Petri Dishes to Two NIR Spectrometers (Data from Ref. (66), Farenhorst, Unpublished Data) R2 Parameter

SEP

Bias

RPD

RER

Landscape A (n=313)

R2

SEP

Bias

RPD

RER

Landscape B (n=295)

Foss NIR Systems 6500 spectrometer 2,4-D

0.92

0.39

-3.07E-07

3.96

17.24

0.86

0.34

-9.79E-06

2.53

11.40

Atrazine

0.93

1.93

-2.38E-05

4.18

16.58

0.91

0.60

-7.08E-07

3.35

14.01

Glyphosate

0.54

16.91

-2.35E-04

1.40

6.97

0.58

60.44

-2.04E-05

1.76

9.01

17β-estradiol

0.92

2.71

8.75E-07

3.78

14.77

0.84

2.42

-4.42E-07

2.53

10.89

SOC%

0.96

0.26

3.89E-07

5.34

18.89

0.91

0.29

-1.86E-06

3.21

12.16

pH

0.67

0.19

2.76E-06

1.70

8.22

0.65

0.39

-2.32E-05

1.71

6.55

Sand%

0.51

10.9

5.48E-06

1.25

6.39

0.70

5.89

-8.01E-06

1.76

10.00

Silt%

0.43

5.51

-6.90E-05

1.35

7.17

0.51

5.92

1.36E-04

1.44

9.03

Clay%

0.48

6.83

4.12E-06

1.05

4.83

0.79

3.28

-6.04E-05

2.13

8.87

45VISNIR Zeiss Corona spectrometer 2,4-D

0.88

0.51

4.84E-06

3.02

14.91

0.82

0.38

-4.51E-09

2.22

10.03

Atrazine

0.85

3.12

1.23E-07

2.42

9.28

0.80

0.87

1.04E-07

2.31

9.68

Glyphosate

0.41

19.19

-1.08E-06

1.34

8.41

0.57

56.22

-7.90E-05

1.90

9.68

17β-estradiol

0.93

2.69

8.06E-05

3.75

15.27

0.81

2.58

-6.99E-07

2.37

10.21

Continued on next page.

In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

R2

SEP

Bias

RPD

RER

R2

SEP

Bias

RPD

RER

Parameter

Landscape A (n=313)

SOC%

0.92

0.37

2.14E-08

3.49

13.04

0.90

0.31

-5.60E-08

2.98

11.26

pH

0.76

0.16

-1.80E-07

2.08

9.18

0.76

0.33

-2.88E-07

2.05

7.83

Sand%

0.55

9.37

7.41E-05

1.83

9.99

0.52

7.31

2.22E-05

1.39

8.06

Silt%

0.49

5.2

-4.03E-05

1.64

10.6

0.41

6.59

3.24E-05

1.29

8.12

Clay%

0.59

5.25

-1.20E-05

1.93

9.09

0.73

3.69

-7.35E-06

1.91

7.89

Landscape B (n=295)

234

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Table 1. (Continued). Calibration Results of 2,4-D, Atrazine, Glyphosate and 17β-Estradiol Kd Values and Selected Soil Properties in Soil Samples Presented in Petri Dishes to Two NIR Spectrometers (Data from Ref. (66), Farenhorst, Unpublished Data)

In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

MIR spectroscopy has been successfully employed to predict sorption of pesticides. Forouzangohar et al. (50) tested this approach on a set of 101 surface soils, collected along a 400 km longitudinal transact from south-eastern South Australia. The soils were characterized using diffuse reflectance infrared Fourier transform (DRIFT) technique. Sorption coefficients of a non-ionic pesticide, diuron [3-(3,4-dichlorophenyl)-1,1-dimethylurea], were measured on these soils by a batch method. The Koc approach was compared against the MIR-PLS approach for prediction of Kd values. They developed and validated the MIR-PLS model by dividing the initial data set into four validation sets. Calibrations were developed using leave-one-out cross-validation which estimates prediction error by removing samples from the calibration set one by one, and predicting them as unknown samples using the remaining calibration samples (81). The model resulted in a coefficient of determination (R2) of 0.69, standard error (SE) of 5.57 and RPD of 1.63. The Koc approach, on the other hand, was found to be inferior with values of R2, SE and RPD of 0.42, 7.26 and 1.25, respectively (Figure 4). To detect spectral outliers, principal component analysis (PCA) was performed on the spectra. It was noted that the significant statistical difference between the two models was mainly due to the outliers detected via PCA. While, the performance of the two models was essentially similar for the rest of the calibration set, this was only true when the best fit Koc was obtained from the measured sorption data on these 101 soils and not taken from the literature. For another herbicide, atrazine, Kookana et al. (82) tested the same approach on a smaller set of surface and subsurface soils (total of 31 samples) and found that compared to the Koc model based on fOC alone, the predictions using MIR-PLS were significantly superior with improved correlation coefficients and lower standard error of estimation.

Figure 4. Measured versus predicted sorption parameter (Kd) using Koc approach (a) and using mid infrared spectroscopy incorporating various soil types (b). Reprinted with permission from Forouzangohar et al., (50). Copyright(2008) American Chemical Society. 235 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

Forouzangohar et al. (83) also made a direct comparison between MIR and visible-NIR (VNIR) for prediction of sorption of diuron on 112 Australian soils (including the 101 soils used in the earlier study). For VNIRS, the spectral data were collected from 400 to 2500 nm with an interval of 2 nm on a NIR 6500 (NIRSystems, Silver Spring, USA) and for MIR spectroscopy, the spectral date were collected for 60 s in the frequency range of 4000 to 500 cm-1 (2500-20000 nm) at 8 cm-1 resolution on Perkin Elmer Spectrum-One FTIR (Perkin Elmer Inc., Mass. USA). A comparison of the performances of VNIR and MIR spectroscopy for predicting both fOC and Kd of diuron was made using PLS regression. In this study, MIR outperformed VNIR for predictions of both of the parameters. MIRPLS model provided a better prediction for diuron Kd values in their calibration set. The key indices R2 and RPD for Kd prediction were 0.8 and 2.3 for the MIR-PLS model and 0.4 and 1.3 for the VNIR-PLS model, respectively.

Concluding Remarks Soil solid phase consists of mineral and organic matters, which are strongly associated with each other and are often difficult to fully separate, despite the common efforts by many workers to isolate one from the other. Both phases contribute towards sorption of pesticides depending on their relative abundance in soil/sediments, their chemistries and the chemical nature of the pesticide molecules. The complex interactions and heterogeneities associated with the soil solid phase contribute to the large degree of variation in the Koc parameter often used to extrapolate the pesticide sorption estimate between soils. One reason for popularity of the Koc model has been ease of its use and absence of any other similar approach that is more comprehensively based on soil properties. Therefore, despite the general recognition in literature of the role of organo-mineral interactions in determining sorption of pesticide, sorption estimates based on a broad range of soil properties have been missing. In this regard, the IR with chemometrics may be worth exploring. The examples presented in this paper show that both MIR and NIR spectroscopy are attractive approaches for rapid assessment of pesticides sorption in soils. In several studies, the assessment by IR-PLS was found to be superior to the Koc model, on the basis of not only the accuracy of prediction, but also being a direct estimation of Kd based on the integrated properties of organic and mineral matter. The Koc approach on the other hand uses an intervening parameter, which considers only one soil property i.e. foc. Furthermore, IR-PLS approach is attractive not only because it allows integration of a range of soil properties (amount and nature of organic carbon as well as mineral matter) in Kd estimation, but also being a relatively high-throughput, cost-effective and easy-to-use technique. While the IR-PLS approach is promising, its broad-based applicability is yet to be demonstrated fully. Further work is needed to assess the transferability of the sorption data predicted by this approach between soils from different regions. 236 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Acknowledgments RK acknowledges the contributions of several students and colleagues, including Riaz Ahmad, Mohsen Forouzangohar, Ahmad Ahangar, Les Janik, Ronald Smernik, for their research over the years, leading to this chapter. Thanks are also due to the anonymous reviewers for their constructive comments on the manuscript.

References

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

1.

2.

3. 4.

5. 6.

7.

8. 9. 10. 11. 12.

13. 14. 15. 16. 17.

Koskinen, W. C.; Harper, S. S. In Pesticides in the Soil Environment: Processes, Impacts and Modeling; Cheng, H. H.; Ed.; Soil Science Society of America: Madison, WI, 1990; pp 51−78. Pignatello, J. J. Interactions of anthropogenic organic chemicals with natural organic matter and black carbon in environmental particles. In BiophysicalChemical Processes of Anthropogenic Organic Compounds in Environmental Systems; Xing, B.Senesi, N., Hunag, P. M., Eds.; Wiley: New York, 2011; pp 3−50. Bollag, J. M.; Myers, C. J.; Minard, R. D. Sci. Total Environ. 1992, 123/124, 205–217. Weber, J. B.; Wilkerson, G. G.; Linker, H. M.; Wilcut, J. W.; Leidy, R. B.; Senseman, S.; Witt, W. W.; Barrett, M.; Vencill, W. K.; Shaw, D. R.; Mueller, T. C.; Miller, D. K.; Becke, B. J.; Talbert, R. E.; Peeper, T. F. Weed Sci. 2000, 48, 75–88. Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Water Res. 1979, 13, 241–248. Hance, R. J. Adsorption and bioavailability. In Environmental Chemistry of Herbicides; Grover, R., Ed.; CRC Press: Boca Raton, Florida, 1988; Vol. 1 pp 1– 20. Theng, B. K. G. Clay-humic interactions and soil aggregate stability. In Soil Structure and Aggregate Stability; Rengasamy , P., Ed.; Seminar Proceedings; Institute of Irrigation and Salinity Research: Tatura, Australia, 1987; pp 32−73. Grathwohl, P. Environ. Sci. Technol. 1990, 24, 1687–1693. Kögel-Knabner, I. Soil Biochem. 1993, 8, 101–105. Karapanagioti, H. K.; Kleineidam, S.; Sabatini, D. A.; Grathwhol, P.; Ligouis, B. Environ. Sci. Technol. 2000, 34, 406–414. Krosshavn, M.; Southon, T. E.; Steinnes, E. J. Soil Sci. 1992, 43, 485–493. Stevenson, F. J. Organic matter reactions involving pesticides in soil. In Humus Chemistry, Genesis, Composition, Reactions, 2nd ed.; John Wiley and Sons, Inc.: New York, 1994; pp 453−471. Inbar, Y.; Chen, Y.; Hadar, Y. Soil Sci. Soc. Am. J. 1989, 53, 1695–1701. Lessa, A. S. N.; Anderson, D. W.; Chatson, B. Plant Soil 1996, 184, 207–217. Baldock, J. A.; Oades, J. M.; Nelson, P. N.; Skene, T. M.; Golchine, A.; Clark, P. Aust. J. Soil Res. 1997, 35, 1061–1083. Schmidt, M. W. I.; Noack, A. G. Global Biogeochem. Cycles 2000, 14, 777–794. Lehmann, J. Nature 2007, 447, 143–144. 237

In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

18. 19. 20. 21. 22.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

43.

Yang, Y. N.; Sheng, G. Y. Environ. Sci. Technol. 2003, 37, 3635–3639. Cornelissen, G.; Gustafsson, O. Environ. Sci. Technol. 2004, 38, 148–155. Kookana, R. S. Aust. J. Soil Res. 2010, 48, 627–637. Kookana, R. S.; Sarmah, A. K.; van Zwieten, L.; Krull, E.; Singh, B. Adv. Agron. 2011, 112, 103–143. Indraratne, S. P.; Farenhorst, A.; Goh, T. B. J. Environ. Sci. Health, Part B 2008, 43, 21–26. Weber, J. B. Soil Sci. Soc. Am. Proc. 1970, 34, 401–404. Kookana, R. S.; Baskaran, S.; Naidu, R. Aust. J. Soil Res. 1998, 36, 715–764. Li, H.; Sheng, G.; Teppen, B. J.; Johnston, C. T.; Boyd, S. A. Soil Sci. Soc. Am. J. 2003, 67, 122–131. Bailey, G. W.; White, J. L. Residue Rev. 1970, 32, 29–92. Quafoku, N. P.; Van Ranst, E.; Noble, A.; Baert, G. Adv. Agron. 2004, 84, 159–215. Regitano, J. B.; Alleoni, L. R. F.; Vidal-Torrado, P.; Casagrande, J. C.; Tornissielo, V. L. J. Environ. Qual. 2000, 29, 894–900. Sarmah, A. K.; Kookana, R. S.; Alston, A. M. Aust. J. Agric. Res. 1998, 49, 775–790. Valverde-Garcia, A.; Gonzalez-Pradas, E.; Villafranca-Sanchez, M.; Del Rey Bueno, F.; Garcia-Rodriguez, A. Soil Sci. Soc. Am. J. 1988, 52, 1571–1574. Khan, Z. A.; Misra, B. M.; Raghu, K. J. Environ. Sci. Health 1996, 31, 1015–1027. Gaultier, J.; Farenhorst, A.; Kim, S. M.; Saiyed, I.; Cessna, A. J.; Glozier, N. E. Wetlands 2009, 29, 837–844. Dios-Cancela, G.; Romero, T. E.; Sánchez-Rasero, F. Soil Sci. 1990, 150, 836–843. Fruhstorfer, P.; Schneider, R. J.; Niessner, R. Sci. Total Environ. 1993, 138, 317–328. Berglöf, T.; Koskinen, W. C.; Brücher, J.; Kylin, H. J. Agric. Food Chem. 2000, 48, 3718–3721. Berglöf, T.; Koskinen, W. C.; Duffy, M. J.; Norberg, K. A.; Kylin, H. J. Agric. Food Chem. 2003, 51, 3598–3603. Roy, C.; Gailllardon, P.; Montfort, F. Pest Manage. Sci. 2000, 56, 795–803. Ahmad, R.; Katou, H.; Kookana, R. S. J. Environ. Qual. 2005, 34, 1045–1054. Weber, J. B.; Wilkerson, G. G.; Reinhardt, C. F. Chemosphere 2004, 55, 157–166. Ahmad, R.; Rahman, A. J. Agric. Food Chem. 2009, 57, 10866–10875. Hornsby, A. G.; Wauchope, R. D.; Herner, A. E. Pesticide Properties in the Environment; Springer Verlag: New York, 1996. Wauchope, R. D.; Yeh, S.; Linders, J. B. H. J.; Kloskowski, R.; Tanaka, K.; Rubin, B.; Katamaya, A.; Kordel, W.; Gerstl, Z.; Lane, M.; Unsworth, J. B. Pest. Manage. Sci. 2002, 58, 419–445. Sabljic A.; Nakagawa, Y. In Non-First Order Degradation and TimeDependent Sorption of Organic Chemicals in Soil; Chen, W., Cheplick, M., Reinken, G., Jones, R. L. Eds.; ACS Symposium Series 1174; American Chemical Society: Washington, DC, 2014; pp 85−118. 238

In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

44. Sabljic, A.; Gusten, H.; Verhaar, H.; Hermens, J. Chemosphere 1995, 31, 4489–4514. 45. Chiou, C. T.; Peters, L. J.; Freed, V. H. Science 1979, 206, 831–832. 46. Hamaker, J. W.; Thompson, J. M. In Organic Chemicals in the Soil Environment; Goring, C. A. I., Hamaker, J. W., Eds.; Marcel Dekker Inc.: New York, 1972; pp 49−143. 47. Mingelgrin, U.; Gerstl, Z. J. Environ. Qual. 1983, 12, 1–11. 48. Gerstl, Z. J. Contam. Hydrol. 1990, 6, 357–375. 49. Ahmad, R.; Kookana, R. S.; Alston, A. M.; Skjemstad, J. O. Environ. Sci. Technol. 2001, 35, 878–884. 50. Forouzangohar, M.; Kookana, R. S.; Forrester, S. T.; Smernik, R. J.; Chittleborough, D. J. Environ. Sci. Technol. 2008, 42, 3283–3288. 51. Ahmad, R.; Kookana, R. S. In Rational Environmental Management of Agrochemicals; Kennedy, I., Solomon, K. R., Gee, S. J., Crossan, A. N., Wang, S., Sánchez-Bao, F., Eds.; ACS Symposium Series 966; American Chemical Society: Washington, DC, 2007; pp 100−119. 52. Laird, D. A.; Barriuso, E.; Dowdy, R. H.; Koskinen, W. C. Soil Sci. Soc. Am. J. 1992, 56, 62–67. 53. Haderlien, S. B.; Schwarzenbach, R. P. Environ. Sci. Technol. 1993, 27, 316–322. 54. Singh, N.; Kookana, R. S. J. Environ. Sci. Health, Part B. 2009, 44, 214–219. 55. Ahangar, A. G.; Smernik, R. J.; Kookana, R. S.; Chittleborough, D. J. Chemosphere 2008, 70, 1153–1160. 56. Boyd, S. A.; Johnston, C. T.; Laird, D. A.; Teppen, B. J.; Li, H. Comprehensive Study of Organic Contaminant Adsorption by Clays: Methodologies, mechanisms and environmental implications. In Biophysical-Chemical Processes of Anthropogenic Organic Compounds in Environmental Systems; Xing, B.Senesi, N., Hunag, P. M., Eds.; Wiley: New York, 2011; pp 51−72. 57. Simpson, M. J.; Simpson, A. J. The role of organic matter-mineral interactions in the sorption of organic contaminants. In BiophysicalChemical Processes of Anthropogenic Organic Compounds in Environmental Systems; Xing, B.Senesi, N., Hunag, P. M., Eds.; Wiley: New York, 2011; pp 73−90. 58. Farenhorst, A.; McQueen, D. A. R.; Saiyed, I.; Hilderbrand, C.; Li, S.; Lobb, D. A.; Messing, P.; Schumacher, T. E.; Papiernik, S. K.; Lindstrom, M. J. Geoderma 2009, 150, 267–277. 59. Kleber, M.; Sollins, P.; Sutton, R. Biogeochemistry 2007, 85, 9–24. 60. Janik, L. J.; Skjemstad, J. O. Aust. J. Soil Res. 1995, 33, 637–650. 61. Williams, P. Near-infrared Technology - Getting the best out of light, A short course in the practical implementation of near-infrared spectroscopy for the user, 5.4 ed.; PDK Grain, PDK Projects, Inc.: Winnepeg, MB, 2009; 182 pp. 62. Martin, P. D.; Malley, D. F.; Manning, G.; Fuller, L. Can. J. Soil Sci. 2002, 82, 413–422. 63. Chang, C. W.; Laird, D. A. Soil Sci. 2002, 167, 110–116. 64. Dunn, B. W.; Batten, G. D.; Beecher, H. G.; Ciavarella, S. Aust. J. Exp. Agric. 2002, 24, 607–614. 239 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.

Downloaded by DUKE UNIV on November 11, 2014 | http://pubs.acs.org Publication Date (Web): October 28, 2014 | doi: 10.1021/bk-2014-1174.ch012

65. Malley, D. F.; Martin, P. D.; Ben-Dor, E. Application in analysis in soils. In Near-Infrared Spectroscopy in Agriculture; Roberts, C. A., Workman, J., Jr., Reeves, J. B., III, Eds.; Agronomy Monography No. 44; American Society of Agronomy, Crop Sciences Society of America, Soil Science Society of America: Madison, WI, 2004; Chapter 26, pp 729−784. 66. Singh, B; Malley, D. F.; Farenhorst, A.; Williams, P. J. Agric. Food Chem. 2012, 60, 9948–9953. 67. McCarty, G. W.; Reeves, J. B., III; Reeves, V. B.; Follett, R. F.; Kimble, J. M. Soil Sci. Soc. Am. J. 2002, 66, 640–646. 68. Reeves, J. B., III; McCarty, G. W.; Mimmo, T. Environ. Pollut. 2002, 116, S277–S284. 69. Cozzolino, D.; Morón, A. Soil Tillage Res. 2006, 85, 78–85. 70. Terhoeven-Urselmans, T.; Kerstin, M.; Helfrich, M.; Flessa, H.; Ludwig, B. J. Plant Nutr. Soil. Sci. 2006, 169, 168–174. 71. Malley, D. F.; Martin, P. D.; McClintock, L. M.; Yesmin, L.; Eilers, R. G.; Haluschak, P. Feasibility of analyzing archived Canadian prairie agricultural soils by near infrared reflectance spectroscopy. In Near-Infrared Spectroscopy; Davies, A. M. C., Giangiacomo, R., Eds.; Proceedings of the 9th International Conference; NIR Publications: Norwich, U.K., 2000; pp 579−585. 72. Sorensen, L. K.; Dalsgaard, S. Soil Sci. Soc. Am. J. 2005, 69, 159–167. 73. Janik, L. J.; Skjemstad, J. O.; Raven, M. D. Aust. J. Soil Res. 1995, 33, 621–636. 74. Zimmermann, M.; Leifeld, J.; Fuhrer, J. Soil Biol. Biochem. 2007, 39, 224–231. 75. Janik, L. J.; Skjemstad, J. O.; Shepherd, K. D.; Spouncer, L. Aust. J. Soil Res. 2007, 45, 73–81. 76. Bornemann, L.; Welp, G.; Brodowski, S.; Rodionov, A.; Amelung, W. Org. Geochem. 2008, 39, 1537–1544. 77. Bengtsson, S.; Berglof, T.; Kylin, H. Bull. Environ. Contam. Toxicol. 2007, 78, 295–298. 78. Singh, B.; Farenhorst, A.; Malley, D. NIR News 2010, 21, 8–10. 79. Farenhorst, A; Papiernik, S. K.; Saiyed, I.; Messing, P.; Stephens, K. D.; Schumacher, J. A.; Lobb, D. A.; Li, S.; Lindstrom, M. J.; Schumacher, T. E. J. Environ. Qual. 2008, 37, 1201–1208. 80. Gerritse, R. G.; Beltran, J.; Hernandez, F. Aust. J. Soil Res. 1996, 24, 599–607. 81. Naes, T.; Isaksson, T.; Fearn, T.; Davies, T. A user-friendly guide to multivariate calibration and classification; NIR Publications: Chichester, 2002. 82. Kookana, R. S.; Janik, L. J.; Forouzangohar, M.; Forrester, S. T. J. Agric. Food Chem. 2008, 56, 3208–3213. 83. Forouzangohar, M.; Cozzolino, D.; Kookana, R. S.; Smernik, R. J.; Forrester, S. T.; Chittleborough, D. J. Environ. Sci. Technol. 2009, 43, 4049–4055.

240 In Non-First Order Degradation and Time-Dependent Sorption of Organic Chemicals in Soil; Kookana, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2014.